The data gives estimates of earnings based on the Labour Force Survey (LFS). Annual figures for the hourly pay of ethnic groups are provided here by taking a four-quarter average of estimates from the LFS.
This analysis uses the median to report average hourly pay; mean hourly pay figures are available in the download file. For each ethnic group, if every respondent was lined up in the order of their hourly pay, the median would be the hourly pay of the person in the middle. The mean is worked out by adding up the hourly pay of all respondents and dividing by the number of respondents.
The choice of the 'mean' as a measure of average hourly pay is affected by ‘extreme’ values which may not reflect some of the lower paid people in each ethnic group. ONS has taken steps to reduce this impact particularly where values over £100 may have been given in error. Using the 'median' as a measure of average pay mitigates against extreme values. However, as with the mean, it is less reliable for ethnic groups where there were small numbers of respondents.
The data on employees’ earnings captured by the LFS is thought to be of a lower quality than other sources, such as the Annual Survey of Hours and Earning (ASHE) and the Average Weekly Earnings survey (AWE). This is because employees report the information themselves rather than their employers. However, this analysis uses LFS data because neither ASHE nor AWE collect data on employees’ ethnicity.
The LFS has been held continuously since 1992. It is designed to produce nationally representative results for any 3-month period.
The sample consists of approximately 40,000 UK households and 100,000 individuals per quarter. Respondents are interviewed 5 times at 3-monthly intervals and 20% of the sample is replaced every quarter. Respondents are only asked about their pay in the first and last of these interviews. The response rate for October to December 2017 was 43%.
All surveys carry the risk of biased results if some people are less likely to respond than others. To compensate for this, the responses to this survey have been weighted so they better reflect characteristics of the target population. The weighting scheme was designed so that the sample reflects the target population's age and sex profile, as well as its geographic spread (region and local authority).
Suppression rules and disclosure control
Sample sizes of less than 30 have been suppressed.
Results above £100 an hour have been excluded because they affect the quality of the data.
‘Suppression’ means these figures have not been included in the data, to protect confidentiality and because the numbers involved are too small to draw any reliable conclusions.
Figures have been rounded to the nearest pence.Quality and methodology information